Do pixels[] follow the same colour model as fill()?

st3f

I'm writing some code that generates a bitmap, pixel by pixel with some partial transparency. The pixels that are partially transparent don't show the colour I expect. Here's an example to illustrate the problem:

background(0,0,0)
size(100,100)
noStroke()
c=color(0x7F7F0000)
fill(c)
rect(50,50,50,50)
loadPixels()
for y in range(50):
    for x in range(50):
        pixels[x+width*y]=c
updatePixels()

I'd expect this to draw two identical squares. The square drawn by 'rect' is as I would expect (dull red). The square drawn by 'pixels' is unexpectedly bright. Is this expected behaviour? And if so, how should I modify my code?

st3f

To show the problem, I added a 'saveFrame()' to the end of my sample code. This generated a third colour which I have screen-shotted next to the output of the sample code.

For the record, since I forgot to include this in the original message, this is on Processing 3.0.1 on OS X 10.11.1

quark

If the fill color is semi-transparent then when you used fill the resulting colour is a merger of the background and fill colour.

When you use pixels you are over-writing the background colour. In this case the alpha value has no effect because the main sketch canvas does not support transparency,

st3f

Thanks, quark. I've switched to using createGraphics, drawing to that, and saving from there. Everything now works as expected. I can ever use image to pop my work onto the canvas after it's complete. Thanks again for a speedy response.